This paper presents some early research on developing a representation for buildings that has the capacity to integrate a number of disparate design domains. This representation is developed based on bond graphs. Its goal is to represent not only steady state topologies of spaces in buildings, but also building dynamics, people and other flows within those spaces.
| Authors | John S. Gero and Jerry Jen-Hung Tsai |
|---|---|
| Year | 2004 |
| Venue | Proceedings of the 9th Conference on Computer Aided Architectural Design Research in Asia (CAADRIA) |
| DOI | 10.52842/CONF.CAADRIA.2004.373 |
| Source Database | Seed (CumInCAD) |
| Bridge-to-GNN | Category Seed |
| GNN Architecture | Not yet extracted from PDF |
| Graph Encoding | Not yet extracted from PDF |
| AEC Task | Not yet extracted from PDF |
| Cohort | Early Infrastructure (2020–2021) |
| Dataset Size | 12 buildings (heuristic — verify) |
|---|---|
| Key Finding | some early research on developing a representation for buildings that has the capacity to integrate a number of disparate design domains (heuristic — verify) |
| Dataset stats found | 12 buildings |
|---|---|
| Quality (heuristic, 6-flag) | 0/6 · rigor: Low Code/Data: —Benchmark: —Baseline: —CV/Split: —Ablation: —Reproducible: — |
Part of the GML/GNN in AEC Systematic Review (PRISMA 2020) — 112 papers, 2020–2026